Azure-docs: Character limits for key phrase extraction

[alathrop]

I am using service level S1 indexing an Excel spreadsheet in blob storage and get the following error:
"message": "Skill #2: Document was truncated to 50000 characters before KeyPhrase extractionrnSkill #1: An error occurred when attempting to obtain entities from portions of input textrn"

Is the truncation to 50,000 characters a limit just for key phrase extraction? I'm assuming this is different that the 4 million character limit for documents in blob storage that is listed on this documentation page. Is the character limit for key phrase extraction different for different service levels? Is this documented?

---

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

• ID: 05e60837-26b9-1f5a-7a44-88ae8c9a7bc3
• Version Independent ID: eaa6fabe-3d51-d626-3a51-d2e46efca454
• Content: Service limits for tiers and skus - Azure Search
• Content Source: articles/search/search-limits-quotas-capacity.md
• Service: search
• GitHub Login: @HeidiSteen
• Microsoft Alias: heidist

[Mike-Ubezzi-MSFT]

@alathrop Thank you for this detailed feedback. We are investigating and will update you when we have accurate information to respond with.

[HeidiSteen]

Hi @alathrop, it's a limit on the key phrase extraction skill: https://docs.microsoft.com/en-us/azure/search/cognitive-search-skill-keyphrases#data-limits

Cognitive skills limits are not documented in service limits, but I'll take an action item to include a section that summarizes them.

[AjayKumar-MSFT]

@alathrop, We will close this out, but if you feel you need more information please just let us know.
@HeidiSteen, Thank you!